ctubbsii commented on PR #2967:
URL: https://github.com/apache/accumulo/pull/2967#issuecomment-1262682666

   > > On the other hand, when a change is submitted, nothing prevents another 
change coming along right behind and changing things to another value. Setting 
a property is not a guarantee that it would not be changed when it is used - so 
this change would not impact that.
   > 
   > Yeah that is how I see it. If there is a good change followed by a 
problematic change, the time between those could be zero,3 seconds, or 4 
months. Throwing a concurrent modification exception offers no reliable 
solution to this. Using conditional updates does offer a possible solution if 
all updates use them because some problematic changes may not be made.
   
   I agree with that. I'm not necessarily suggesting we need to try to detect 
that. I'm more concerned that we're adding extra code that goes out of our way 
to pretend it didn't happen and hide the fact from the user when we did detect 
it. I don't see a reason to do that. I don't think we should try to detect the 
case where it could happen silently... I'm just saying we shouldn't be the ones 
perpetuating the silence when it could have otherwise been made known to the 
user that it did happen.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to